A procedure to diagnose the faults in mining equipment used in long wall coal mining is proposed. The procedure is based on creating a system model and comparing outputs from the system model with outputs from the operational system. A fault in the operational system is detected as a residual error between the operational system and the model. The comparisons are made using a suboptimal fading extended Kalman filter combined with a modified Bayes estimation algorithm. A simulated case study is used to demonstrate the efficiency of the procedure. The simulation comprises a dynamic model of a shunt DC motor driven system and the corresponding fault state observer is used to monitor the operating conditions on line.
展开▼